Long-term Central Waikato summer-autumn rainfall and pasture growth trends. Are conditions for pasture growth changing over time?
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
127
Long-term Central Waikato summer-autumn rainfall and pasture
growth trends. Are conditions for pasture growth changing over time?
Chris B GLASSEY1,*, R. Grant WILLS4, Mike B DODD2, Kieran S McCAHON1 and David F CHAPMAN3
1DairyNZ, Private Bag 3221, Hamilton 3240, New Zealand
2AgResearch, 11 Dairy Farm Road, Palmerston North 4442, New Zealand
3DairyNZ, 24 Millpond Lane, Lincoln 7608, New Zealand
493 Paratu Road, RD1, Matamata 3475, New Zealand
*Corresponding author: chris.glassey@dairynz.co.nz
Abstract farmers are already concerned that summers and
Farmers in the upper North Island are concerned autumns in recent years have been more variable, and
about the poor productivity of their perennial ryegrass generally drier and warmer than normal, resulting
pastures beyond 3 years and suggest this is linked to a in more challenging conditions for pasture growth.
trend towards drier conditions for pasture growth during Farmers are also noticing increasing costs and effort
summer and autumn. To explore how conditions for for maintaining their pasture base and this is reflected
pasture growth and survival have changed, trends in in costs recorded by dairy industry economic surveys
rainfall (measured and interpolated; c. 1954 to 2020), (Dodd et al. 2018).
frequency of soil moisture deficit stress on pasture Evidence to support these farmer observations is
growth (1972-2020), and pasture growth rates (simulated sparse. Glassey (2011) reported a decline in mean
1977-2020, and measured c. 1979-2020) were compiled December to April rainfall at Ruakura of around 40
for summer-autumn months (November to April) mm per decade since 1979, an increase in variability
for two Waikato locations: DairyNZ Ruakura/Scott of rainfall with coefficient of variation (CV) increasing
Farm near Hamilton; and a commercial farm at Paratu from 19% in the 1980s and early 1990s to 29% in
Road between Morrinsville and Matamata. Significant the early 2000s, and a negative relationship between
negative linear trends were observed for interpolated rainfall and cumulative pasture growth (-85 kg dry
rainfall (Paratu Road only), total stress days (both sites) matter (DM)/ha per 10 mm-reduction in total rainfall).
and simulated pasture growth (both sites). No significant These trends, and their potential effects on other pasture
trend in measured pasture growth was observed for either performance factors such as pest outbreaks (Ferguson et
site, suggesting adaptive management has cushioned the al. 2019), result in increased risk to pasture persistence
physical effects of harsher climatic conditions. The suite which is reflected in farmer observations (Tozer et al.
of adaptive management practices adopted on the Paratu 2011) and confirmed in trial data (Lee et al. 2017).
Road farm is described. Analysis of decadal patterns This between-year variability for central Waikato
indicated greater variability in rainfall in the decade summers, which are often dry, means that recent dry
2011-2020 (coefficient of variation ranging from 32% years need to be placed in an historical context to
to 36% depending on site and data source, versus ~ 22% help identify the right adaptation strategies for the
in earlier decades) and a significantly higher number future dairy forage base in the region. Since 2010,
of stress days in the decade 2011-2020 compared with measurement of pasture growth at the site reported
the 1970s and 1980s at Paratu Road. Further adaptation by Glassey (2011) has continued. A long-term
strategies will be needed to mitigate these most-recent pasture growth dataset was also available from a
trends, should they continue as predicted by climate local commercial dairy farm. In addition, models are
change forecasts. available for evaluating long-term trends for summer-
autumn pasture growth for Waikato locations based on
Keywords: perennial ryegrass, persistence, summer profile available water (PAW, as defined in Overseer™)
moisture deficit trends, Rezare Pasture Growth and actual climate data, e.g., the Rezare Pasture Growth
Forecaster, Virtual Climate Network Forecaster (Ogle 2015). Simulation of pasture growth
enables an analysis independent of other farm system
Introduction influences (e.g., fertiliser, supplement use and grazing
Future climate change projections for New Zealand management) which may help elicit the extent to which
point to temperature increases, particularly in summer- those farm systems management factors may already be
autumn; more frequent hot days; precipitation decreases off-setting climate-driven trends in pasture production.
in northern and eastern regions; and increased drought This is useful for helping gauge what capacity exists for
severity (Mullan et al. 2018). In central Waikato, adaptation (albeit likely to add to costs of production)
ISSN 0118-8581 (Print) ISSN 2463-4751 (Online) https://doi.org/10.33584/rps.17.2021.3456
128 Resilient Pastures – Grassland Research and Practice Series 17: XX-YY (2021)
versus the need to develop further adaptation measures. grid node nearest Scott Farm (latitude 37.4606 south,
The objective of this study was to investigate long- longitude 175.2200 east), were examined for long-
term rainfall trends and the combined impact of rainfall term rainfall and soil moisture deficit trends over six
plus soil moisture stress on plants over the summer- months from November to April, and rainfall for each
autumn months and how these have affected conditions individual month within that period.
for summer-autumn growth of perennial ryegrass-based
pastures in the central Waikato region, particularly over Net herbage accumulation (HA)
the past 4 decades. Simulated net HA
Both measured and modelled pasture growth were Pasture growth modelling was used to translate
used for this purpose, noting that data for measured long-term weather patterns into potential pasture
pasture growth over periods greater than 3 to 5 years are production, independent of the effects of changes in
sparse. We also note that 4 decades is still a relatively farm management (e.g., stocking rate, soil fertility).
short time from which to draw firm conclusions The Rezare Pasture Growth Forecaster (Ogle 2015)
given the evidence for decadal-scale climate patterns was used to simulate daily pasture growth potential
(Interdecadal Pacific Oscillation, Folland et al. 2002). for the two sites from November 1977 to May 2020
We therefore put forward observations rather than firm (based on the availability of daily weather data from
conclusions and, in so doing, highlight the critical data/ the VCN). The model utilises information on farm
knowledge gaps that should be addressed to provide type, geographical location, plant available water, daily
clearer messages for farmers about their exposure to weather, fertiliser and irrigation inputs to generate daily
risk and associated adaptation options. changes in pasture biomass (Romera et al. 2009, 2013).
Pasture growth potential in the model is influenced
Materials and Methods strongly by temperature and moisture growth-limiting
Analysis of long-term climate and pasture growth factors that are combined into a single stress index.
trends focussed on two central Waikato locations, This index typically accumulates during summer when
encompassing two research farms east of Hamilton pastures are moisture-limited and temperatures are
(Ruakura No. 2 dairy, and DairyNZ’s Scott Farm at high, then declines during autumn as moisture and
Newstead) and a commercial dairy farm 21 km south of temperature stresses are alleviated. Predicted daily
Morrinsville (Paratu Road). This analysis included both net HA (kg DM/ha/day, Hodgson 1979) and the daily
actual and modelled pasture growth data from various value of the stress index, were compiled for each of
sources as outlined below. the 43 growing seasons to indicate long-term trends
in climatically induced stress and annual pasture
Weather and climate data production, as per Dodd et al. (2018).
Daily rainfall records from the Ruakura climate station For the two locations, the site-specific input data
(National Institute of Water and Atmospheric Research for the model included the daily weather file from the
(NIWA) 26177 EWS, latitude 37.7757 south, longitude nearest VCN node and the soil profile available water
175.3051 east) were aggregated into monthly rainfall (PAW) from S-Map (https://smap.landcareresearch.
totals for November to April (incl.) from 1954 to 2020. co.nz/) for the predominant soil type on each farm. For
The 6 months from November to April were chosen as Scott Farm PAW was set at 198 mm (Matangi silt loam)
the months where it was most likely that the interaction and for Paratu Road the PAW was set at 133 mm for (Te
between rainfall and potential evapotranspiration Rahu silt loam).
creates soil moisture deficits that affect pasture growth. Using the VCN data we calculated a daily soil
This was called the summer-autumn period and aligns moisture balance for summer-autumn based on
with the November-January and February-April incoming daily rainfall (mm), minus daily potential
months as defined for the upper North Island for the evapotranspiration (PET, mm), and a fixed available
DairyNZ Forage Value Index (DairyNZ 2021). The water capacity (PAW, the amount of water in the soil
variability (CV%) of rainfall for November to April ‘reservoir’ that plants can use). From this we determined
for the Ruakura climate station was calculated for each the number of “stress days” for each year where plant
decade over this period. roots take up water with increasing difficulty and plant
Interpolated climate data were available from the growth is restricted.
Virtual Climate Network (VCN, Tait et al. 2006), which
uses daily NIWA climate station records (NIWA 2020) Measured net HA
to estimate values for a network of sites across New Measured monthly pasture growth rates were collated
Zealand on a 5 km grid. VCN data from 1960-2020 for for summer-autumn (November to April) from two
the grid node nearest to the Paratu Road farm (latitude research sites and one commercial dairy farm.
37.4511 south, longitude 175.3842 east), and another Records from Ruakura No. 2 Dairy (1979-2004) and
Glassey et al., Long-term Central Waikato summer-autumn rainfall and pasture growth trends. Are conditions for pasture growth... 129 then Scott Farm (2004 to 2020) were compiled, the VCN rainfall, soil moisture (stress days), and measured latter adding 10 additional years to a previous analysis and modelled pasture growth (net HA) at each site were (Glassey 2011). These two farms are less than 5 km examined using linear regression, scatterplot smoothing apart and have similar perennial ryegrass/white clover- utilizing locally weighted polynomial regression dominant pastures. They are referred to hereafter as (LOWESS, Cleveland 1979), and box plots by decade. ‘Scott Farm’. The following data were analysed for If required to achieve homogeneity of variance, data between-year variability in summer-autumn net HA. were transformed for analysis. In order to test for equal • Monthly net HA for the ‘control’ farmlet at Ruakura variances between decades, Bonferroni 95% confidence No. 2 Dairy from June 1979 until May 1993. Data intervals, adjusted for multiple comparison, were used. from 11 of 13 years were available. No nitrogen (N) There was no significant indication of unequal variances fertiliser was applied to pastures during this period. for the different decades (Tables 2 and 3). • Monthly net HA for the ‘control’ farmlet at Ruakura Box plot graphs by decade are presented except for No. 2 Dairy from June 1993 until May 2004. During measured and modelled pasture growth due to Paratu this period annual N fertiliser application to pastures Road measured data spanning only 2.6 decades. Where averaged 186 kg N/ha. the slope of the linear regression was significantly • Monthly net HA for the ‘control’/’benchmark’ different from zero, they are reported in the text and farmlet at Scott Farm from June 2004 to May 2020. in tables. Differences were considered significant at During this period, annual N fertiliser applications P
130 Resilient Pastures – Grassland Research and Practice Series 17: XX-YY (2021)
Figure 1 Cumulative summer-autumn (November-April) interpolated rainfall (mm) by decade from VCN climate data sites
representing Scott Farm and Paratu Road. Boxes encompass the middle quartiles, plus the median (solid line). Whiskers
represent upper and lower quartiles.
Figure 1 Cumulative summer-autumn (November-April) interpolated rainfall (mm) by
rainfall among decades (Figure 1), although the CV at Scott Farm (Figure 2), but a significant increase in
decade frominVCN
increased climate decade
the most-recent data sites
(Tablerepresenting
3) at both Scottdays
stress Farm and
in the Paratu
most recentRoad.
decade Boxes
compared with
sites, similar to the pattern in measured rainfall at earlier decades at Paratu Road.
encompass the middle quartiles, plus the median (solid
Ruakura noted in Table 1.
line). Whiskers represent upper and
Linear regression analysis of summer-autumn stress
lowerWhen individual months (November to April) were
quartiles. days by individual months showed significant increases
analysed for trends in interpolated rainfall, there was over time for November, December and January at Scott
a significant decline for November for both locations Farm, and for November and January at Paratu Road
(P=0.021 for Scott Farm; P=0.005 for Paratu Road) with a statistical trend toward an increase in December
Whenbut individual
no significantmonths (November
trends were observed fortotheApril)
other were analysed
at Paratu for trends
Road (Table 4). in interpolated
months at either location (data not presented).
rainfall, there was a significant decline for November for both
Simulated net HAlocations (P=0.021 for Scott
Soil moisture stress days The model predicted a significant linear decline for
Farm; P=0.005
There for Paratu
was a significant Road)
linear butin no
increase significant
stress days trends werepasture
summer-autumn observed forfrom
growth the 1977
othertomonths
2020
from 1972-2020 at both sites (Table 1). There were no for both VCN nodes representing Scott Farm and
at either location
differences among(data notinpresented).
decades calculated stress days Paratu Road (Table 1). However, the linear component
Table 2 Mean summer- autumn rainfall at Ruakura (November-April, mm), 1954-2019 by decade. Includes standard deviation
(SD) from mean and coefficient of variation, CV%.
Soil moisture stress days
Decadal variability Decade
There
was a significant linear increase in stress days from 1972-2020 at both sites (Table 1).
1950 1960 1970 1980 1990 2000 2010 Overall mean
There were no differences among decades in calculated stress days at Scott Farm (Figure 2),
Nov-Apr rainfall Mean 574 556 498 548 489 488 505 516
but a significant increase in stress days in the most recent decade compared with earlier
SD 118 109 105 103 97
119 108 164
CV% 21 20 21 19 20 22 32 23
decades at Paratu Road.Glassey et al., Long-term Central Waikato summer-autumn rainfall and pasture growth trends. Are conditions for pasture growth... 131
225
Figure 2 Cumulative number of moisture stress days in summer-autumn (November-April) calculated by decade from VCN climate
226 data sites representing Scott Farm and Paratu Road. Boxes encompass the middle quartiles, plus the median (solid line).
Whiskers represent upper and lower quartiles.
227 Figure 2 Cumulative number of moisture stress days in summer-autumn (November-April)
228 calculated by decade from VCN climate data sites representing Scott Farm and Paratu Road.
Table 3 Decadal variability (standard deviation, SD; and coefficient of variability, CV%) in interpolated cumulative summer-
229 Boxes encompass the middleinclusive)
autumn (November-April quartiles,
rainfallplus
(mm) the
from median (solid
VCN climate line).
data sites Whiskers
representing Scott represent upper
Farm and Paratu Rd
from 1960-2020.
230 and lower quartiles.
Decadal variability Decade
231
1960 1970 1980 1990 2000 2010
232 Linear regression analysis of summer-autumn stress days by individual months showed
Nov-Apr rainfall Scott Farm SD 129 125 100 97 106 176
233 significant
increases over timeCV%
for November,
23 December
23 and
19 January 20
at Scott Farm,
23 and35for
234 November and January
Paratu Rd at Paratu
SD Road with
118 a statistical
127 trend
96 toward an
91 increase
128 in December
179
CV% 20 22 16 16 25 36
235 at Paratu Road (Table 4).
Table 4 Linear trends (slope) and statistical significance (P-value) by month for summer-autumn moisture stress days at Scott
236
Farm and Paratu Road, 1972-2020.
237 Table 4 Linear trends (slope) and statistical significance (P-value) by month for summer-
Month
238 autumn moisture stress days at Scott Farm and Paratu Road, 1972-2020.
Linear trend by month Nov Dec Jan Feb Mar Apr
Stress days Scott Farm P-value 0.0447 132 Resilient Pastures – Grassland Research and Practice Series 17: XX-YY (2021)
Modelled pasture growth November to April (kg DM/ha)
8
Figure 3 Predicted summer-autumn net HA (November-April) from 1977- 2020 for Scott Farm and Paratu Road. Shaded areas
9 Figure 3 Predicted summer-autumn
are 95% confidence intervals. net HA (November-April) from 1977- 2020 for Scott
0 Farm
onlyand Paratu20%
explained Road. Shaded
(Scott Farm) areas are(Paratu
and 22% 95% confidence intervals.
The variability between years dominated much of
Road) of the overall variation in predicted cumulative the available data making detection of trends difficult.
1 summer-autumn HA. The LOWESS analysis (Figure 3) For example, it was difficult to confirm the trend in
indicated that most of the decline in predicted pasture summer-autumn rainfall at Ruakura over 70 years, as
2 Observed net HA
growth occurred from the mid-1990s to 2020. the linear trend only approached significance with a
3 Mean cumulative net HA for November to April inclusive measured
very low regression during(P=0.075,
co-efficient the Ruakura No.
r2 = 0.048).
Observed net HA We used three independent statistical methods to
4 2 Dairy
Mean and Scott Farm
cumulative net HAsequence
for Novemberwas 9.4 t DM/haexplore
to April ±1.4 trends
SD over 41 the
within years.
data,At Paratu
each with Road,
their own
inclusive measured during the Ruakura No. 2 Dairy and respective advantages and limitations. The chosen time
5 mean cumulative net HA for November to April was
Scott Farm sequence was 9.4 t DM/ha ±1.4 SD over
7.1 t DM/ha ±1.6 SD over 26 years.
period for analysis, and the influence of individual year
6 41 years. At Paratu
Inter-annual Road,inmean
variation net cumulative net HA for
HA was greater data points
at Paratu Roadin (CV=22%)
relation to the compared
time sequencewith
of available
November to April was 7.1 t DM/ha ±1.6 SD over 26 data, can influence the sensitivity of a linear regression.
7 Ruakura/Scott Farm
years. Inter-annual (CV=14%).
variation in net HA There was no
was greater at linear trend
Because of over time in evidence
climatological NHA for foreither site
decadal-scale
Paratu Road (CV=22%) compared with Ruakura/Scott oscillations in weather patterns, we grouped the data
Farm (CV=14%). There was no linear trend over time by decade. This represents a somewhat arbitrary
in NHA for either site (Table 1). 10 decision of which 10 years will constitute a decadal
group. We explored this despite the loss of statistical
Discussion power compared with using individual years for linear
The results suggest that farmers in at least two locations or LOWESS analysis. We present box plots graphs by
in central Waikato are facing a trend of increased decade as they give a much better depiction of any change
frequency of summer-autumn soil moisture deficits in variability than the scattered individual year patterns
while still coping with substantial (and possibly used in the linear regression and LOWESS graphs.
increasing) variability in rainfall between years. The availability of VCN data strengthened our search
Salinger & Porteous (2014) reported a distinct trend for evidence and allowed inclusion of the site at Paratu
towards higher values of drought indices over a 72-year Road with recorded pasture production for 26 years
period, including the region covered in our analysis. (Figure 4). Keeping monthly rainfall records on Paratu258 (Table 1).
Glassey et al., Long-term Central Waikato summer-autumn rainfall and pasture growth trends. Are conditions for pasture growth... 133
259
November to April pasture growth (kg DM/ha)
260
261 Figure 4 Measured summer-autumn net HA (November-April, kg DM/ha) for Ruakura/Scott Farm (1979-2020) and Paratu Road
(1995-2020). Shaded areas are 95% confidence intervals.
262 Figure 4 Measured summer-autumn net HA (November-April, kg DM/ha) for Ruakura/Scott
Road was abandoned many years ago after recognition one location. Long-term sequential records for annual
263 Farm
of the(1979-2020)
large effects and
that Paratu Roadof(1995-2020).
the timing individual Shaded
pasture areas
growth areresearch
from 95% confidence intervals.
sites in Waikato before
rainfall events can have on soil moisture levels and, 1980 were difficult to find although some publications
264 therefore, pasture growth. The VCN data allowed the presented data for shorter periods (e.g., Mcaneney et al.
265 Discussion
calculation of daily soil moisture balances and provided (1982) covered 1953-1966 and Baars (1976) covered
266 information on the frequency of “stress days”, where 1954-1970).
267 The
soil results
moisturesuggest thatpasture
was limiting farmers in at least two locations
growth. in central
For measured Waikato
pasture growtharewe
facing
were a unable
trend
The analyses of VCN data supported farmer concerns to detect any significant trends over time at both
268 ofofincreased
increasinglyfrequency
drier and more of variable
summer-autumn
summers. While soil moisture deficits4).while
locations (Figure still coping
This contrasts with
with the simulated
the VCN data box plot for rainfall by decade for both summer-autumn pasture growth (Figure 3). Simulated
269 substantial
sites showed(and possiblytrend
no significant increasing) variability
at Scott farm there in rainfall between
cumulative NHA for years. Salinger fell
November-April & broadly
was a linear trend (P=0.06) towards reduced summer- within the range measured at Scott Farm (5000-12000
270 Porteous (2014) reported a distinct trend towards higher values of drought indices over a 72-
autumn rainfall, and for Paratu Road there was a kg DM/ha). For Paratu Road the model predicted
271 significant (P=0.03) linear decline in summer-autumn
year period, including the region covered in our analysis. higher pasture growth (by ~2000 kg DM/ha on average)
rainfall over the past 6 decades. For Paratu Road the than the measured data. A possible explanation for the
272 The variability
boxplot by decade between years
also showed dominated
a significant much ofdissimilarity
increase the available in thedata making
long-term trenddetection of
between measured
in soil moisture deficit days for the most recent decade and simulated data sets is that the pasture growth model
273 trends
(Figuredifficult.
2). For example, it was difficult to confirm uses onlythelocaltrend in summer-autumn
environmental rainfall
conditions (weather and
Additional analysis of VCN data for each location by soil) in its predictions, whereas on-farm measurements
274 atmonth
Ruakura over 70 years, as the linear trend onlyofapproached
showed November to be the only month with a
significance with a very low
pasture growth will also be influenced up or down by
275 consistent and
regression significant decline
co-efficient (P=0.075, r2 = over
in rainfall time
0.048). other factors, such as reducing feed demand by culling,
and this resulted in significant increases in soil moisture N applications, supplement use, rotation length and
276 Wefrom
stress days usedNovember
three independent statistical
to January (Table 3). methods
managingto explore trends(Reynolds
grazing intensity within the data, each
2013).
Linking trends in climate data to measured pasture The significant decline in summer-autumn pasture
277 with their own
production respective
is difficult becauseadvantages
of the lack and limitations.
of actual growthThe chosen
predicted by time period
the model for analysis,
perhaps confirms that
long-term pasture growth measurements from any the trends found for declining summer-autumn rainfall,
278 and the influence of individual year data points in relation to the time sequence of available
11134 Resilient Pastures – Grassland Research and Practice Series 17: XX-YY (2021) including November, at both farms are creating stress and threatening their persistence, there is also a challenges for managing feed supply and feed demand cost of ‘doing nothing’ as farmers become trapped in a for Waikato dairy farmers, especially on soil types with cycle of re-grassing and re-cropping (Dodd et al. 2018). lower PAW (e.g., Paratu Road). It is evident that the magnitude of variation appears Conclusions/Practical implications/Relevance to have increased for both sites in the most recent This study suggests that farmers near the locations decade (Tables 1 and 2; Figures 1 and 2). This is examined are likely to have experienced increased consistent with climate projections for increases in variability and frequency of summer moisture deficits, the frequency of extreme conditions (precipitation, which our pasture growth modelling shows is likely to temperature and wind, Mullan et al. 2018). Increased have been accompanied by a declining trend over time variability is a likely contributor to reduced resilience in pasture accumulation rates. of ryegrass pastures (Lee et al. 2017). It also impacts Depending on their farm’s risk profile for ryegrass/ on the choice of farm system through a reduction in clover pasture resilience, farmers in the upper North opportunity to carry forward surplus feed from one Island will need to explore alternatives to perennial season to the next and increases other risk factors such ryegrass to maintain their future home-grown feedbase, as soil physical damage associated with increased crop/ or adapt their pasture management to cope with the pasture establishment method and management. increasing risk of summer moisture deficit. The increased frequency of drier summers in the The integration of VCN data with pasture growth past decade (Figures 2 and 3) will also be influencing models appears to be an opportunity to help farmers farm management changes over time due to the reduced understand how their local climate is behaving and amount of pasture available in drier summers. For should inform farm management decisions that help example, Glassey (2011) reported that 100 mm less them cope with the increasing risk of summer soil rainfall between December and April was associated moisture deficits. with 850 kg DM/ha less pasture grown. Adding more sophisticated analysis such as thermal Farm management practices have been adapted time accumulation, and other statistical methods, could over time at the Paratu Road farm in response to the add increased certainty to these messages. variability in farm-specific summer-autumn pasture growth measurements. These include: ACKNOWLEDGEMENTS • An increase in weed spraying because of more Grant Wills provided 26 years of monthly pasture Setaria pumilia (Poir.) (yellow bristle grass) and growth data for his dairy farm at Paratu Road, Walton. other C4 grasses. Barbara Kuhn-Sherlock (DairyNZ) provided valuable • Increased use of winter-active ryegrass cultivars statistical advice. We are also grateful to the many such as ‘Shogun’ integrated with a summer cropping technicians who have contributed to the collection and programme. storage of monthly pasture growth data over the years • Use of three Herd Homes™ in summer to help at Ruakura No. 2 dairy and Scott Farm. control post-grazing residuals and mitigate heat stress for cows. REFERENCES • Increased imported feed in the last 12 years to Baars JA. 1976. Seasonal distribution of pasture mitigate the variability of pasture production. production in New Zealand. IX Hamilton. New • A change (4 years ago) to 25% of the herd starting Zealand Journal of Experimental Agriculture 4: 157- calving in March, with a winter milk contract. This 161. https://doi.org/10.1080/03015521.1976.10425 matches the measurements of winter pasture growth 862 being much more reliable than in summer. Cleveland WS. 1979. Robust locally weighted • Nitrogen fertiliser use has reduced to about 120 kg regression and smoothing scatterplots. Journal N/ha/yr and is largely confined to spring (or autumn), of American Statistical Association 74: 829−836. when soil moisture is most reliable. https://doi.org/10.1080/01621459.1979.10481038 Some of these changes, such as use of the Herd DairyNZ. 2021. Forage Value Index. The DairyNZ Homes and more imported feed, provide options for FVI Handbook 2021 Edition. 47p. Retrieved 19 tactical management of pastures over late spring- February 2021 from: https://www.dairynz.co.nz/ summer-autumn to, for example, prevent over-grazing media/5794012/forage-value-index-handbook-2021- and support pasture persistence. However, farm system final.pdf. adaptations such as those mentioned above bring Dodd MB, Chapman DF, Ogle G. 2018. Re-grassing additional costs into the farm system. In situations where trends and drivers in the New Zealand dairy industry. variability in summer-autumn growing conditions is Journal of New Zealand Grasslands 80: 177-183. placing perennial ryegrass pastures under significant https://doi.org/10.33584/jnzg.2018.80.342
Glassey et al., Long-term Central Waikato summer-autumn rainfall and pasture growth trends. Are conditions for pasture growth... 135 Ferguson CM, Barratt BIP, Bell N, Goldson SL, the IPCC Fifth Assessment, 2nd Edition. Ministry for Hardwick S, Jackson M, Jackson TA, Phillips the Environment, Wellington, New Zealand. 131 p. CB, Popay AJ, Rennie G, Sinclair S, Townsend R, NIWA. 2020. Virtual Climate Station data and products. Wilson M. 2019. Quantifying the economic cost of Retrieved 16 November 2020 from: https://niwa. invertebrate pests to New Zealand’s pastoral industry. co.nz/climate/our-services/virtual-climate-stations New Zealand Journal of Agricultural Research 62: Ogle G. 2015. Forecasting pasture growth rates 255-315. https://doi.org/10.1080/00288233.2018.14 across New Zealand. Retrieved 16 November 2020 78860 from: https://www.rezare.co.nz/forecasting-pasture- Folland CK, Renwick JA, Salinger MJ, Mullan growth-rates-across-new-zealand/ AB. 2002. Relative influences of the Interdecadal R Core Writing Team. 2013. R: a language and Pacific Oscillation and ENSO on the South environment for statistical computing. R Core Team, Pacific Convergence Zone. Geophysical R Foundation for Statistical Computing, Vienna, Research Letters 29:21-1-21-4. https://doi. Austria. Available at: www.R-project.org org/10.1029/2001GL014201 Reynolds WN. 2013. Addressing on-farm management Glassey CB. 2011. Summer pasture yield variation to enhance pasture productivity and persistence. in a central Waikato location from 1979 to 2010: Proceedings of the New Zealand Grassland implications for pasture persistence. In: Mercer CF. Association 75: 241-244. https://doi.org/10.33584/ Ed. Pasture Persistence Symposium. Grassland jnzg.2013.75.2914 Research and Practice Series 15. Dunedin, New Romera AJ, McCall DG, Lee JM, Agnusdei MG. 2009. Zealand: New Zealand Grassland Association, pp. Improving the McCall herbage growth model. New 15-20. https://doi.org/10.33584/rps.15.2011.3214 Zealand Journal of Agricultural Research 52: 477- Hodgson J. 1979. Nomenclature and definitions in 494. https://doi.org/10.1080/00288230909510529 grazing studies. Grass and Forage Science 34: 11-18. Romera A, Beukes P, Clark D, Clark C, Tait A. 2013. https://doi.org/10.1111/j.1365-2494.1979.tb01442.x Pasture growth model to assist management on dairy Lee JM, Thom ER, Waugh CD, Bell NL, McNeill farms: Testing the concept with farmers. Grassland M.R, Wilson DJ, Chapman DF. 2017. Trajectory Science 59: 20-29. https://doi.org/10.1111/grs.12009 and causes of decline in the botanical composition Salinger MJ, Porteous AS. 2014. New Zealand climate: of dairy-grazed pasture in the Waikato. Journal of patterns of drought 1941/42-2012/13. Weather and New Zealand Grasslands 79: 89-96. https://doi. Climate 34: 2-19. https://doi.org/10.2307/26169741 org/10.33584/jnzg.2017.79.546 Tait A, Henderson R, Turner R, Zheng X. 2006. Thin Macdonald KA, Penno JW. 1998. Management plate smoothing spline interpolation of daily rainfall decision rules to optimise milksolids production on for New Zealand using a climatological rainfall dairy farms. Proceedings of the New Zealand Society surface. International Journal of Climatology 26: of Animal Production 58: 132–135. 2097-2115. http://dx.doi.org/10.1002/joc.1350 Mcaneney KJ, Judd MJ, Weeda WC. 1982. Loss in Tozer KN, Cameron CA, Thom ER. 2011. Pasture monthly pasture production resulting from dryland persistence: farmer observations and field conditions in the Waikato. New Zealand Journal of measurements. In: Mercer CF. Ed. Pasture Agricultural Research 25: 151-156. https://doi.org/1 Persistence Symposium. Grassland Research and 0.1080/00288233.1982.10420904 Practice Series 15. Dunedin, New Zealand: New Mullan B, Sood A, Stuart S, Carey-Smith T. 2018. Zealand Grassland Association, pp. 25-30. https:// Climate Change Projections for New Zealand: doi.org/10.33584/rps.15.2011.3216 Atmosphere Projections Based on Simulations from
You can also read
